Celebrity Look-Alike

A web application that uses facial recognition to find your celebrity doppelgänger! Upload a photo, and the app compares your facial features to a dataset of celebrity images to identify your closest match. Built with Python, Streamlit, and advanced machine learning models, this project offers a fun and interactive way to discover which celebrity you resemble.

Features

• Image Upload: Easily upload a photo (jpg, jpeg, png, or gif format) to find your celebrity look-alike.
• Facial Recognition: Utilizes MTCNN for face detection and FaceNet for facial feature extraction to ensure accurate comparisons.
• Real-Time Results: Displays the closest celebrity match alongside your uploaded image with a similarity score.
• User-Friendly Interface: Built with Streamlit for a seamless and intuitive experience.
• Privacy-Focused: Images are processed locally and not stored after analysis.

How It Works

1. Face Detection: The app uses MTCNN to detect faces in the uploaded image, ensuring only single-face images are processed for accuracy.
2. Feature Extraction: FaceNet extracts a 128-dimensional vector representing the facial features of the detected face.
3. Similarity Matching: The extracted vector is compared to a precomputed dataset of celebrity face embeddings using cosine similarity to find the closest match.
4. Result Display: The app shows your uploaded image side-by-side with the matched celebrity's image and a similarity score.

Demo

Below are examples of the app in action:

User Photo and Celebrity Match

Uploaded photo (left) matched with a celebrity (right) with a similarity score of 0.65.

Installation

To run the Celebrity Look-Alike app locally, follow these steps:

Prerequisites

• Python 3.8 or higher
• venv
• pip package manager
• Git

Steps

1. Clone the Repository:

   git clone https://github.com/mc095/Celebrity-Look-Alike.git
   cd Celebrity-Look-Alike

2. Create a Virtual Environment:

   python -m venv venv
   venv\Scripts\activate

3. Install Dependencies:

   pip install -r requirements.txt

4. Run the App:

   python -m streamlit run app.py

   Open your browser and navigate to http://localhost:8501 to use the app.

Project Structure

• app.py: Main Streamlit application script.
• celeb_images/: Directory for storing celebrity image dataset.
• demo/: Contains demo images for README.
• requirements.txt: List of Python dependencies.

Dependencies

• streamlit: For the web interface.
• facenet-pytorch: For face detection and feature extraction.
• opencv-python: For image processing.
• numpy, pandas: For data handling and similarity calculations.
• matplotlib: For visualizing results.

Dataset

The app uses a dataset of celebrity images (e.g., from IMDB or Wikipedia). Ensure the dataset is formatted as follows:

• A folder containing subfolders for each celebrity, with images named appropriately.

Usage

1. Open the app in your browser.
2. Upload a clear photo of your face (ensure only one face is visible, no hats or obstructions).
3. Click "Find My Celebrity Look-Alike" to process the image.
4. View the result, which includes your photo, the matched celebrity's photo, and the similarity score.

Limitations

• Single Face Requirement: The app works best with images containing only one face. Group photos may lead to inaccurate results.
• Image Quality: Low-resolution or blurry images may reduce matching accuracy.
• Dataset Dependency: The quality and size of the celebrity dataset affect the results. Expand the dataset for better matches.

Contributing

Contributions are welcome! To contribute:

1. Fork the repository.
2. Create a new branch (git checkout -b feature-branch).
3. Make your changes and commit (git commit -m "Add feature").
4. Push to the branch (git push origin feature-branch).
5. Open a pull request.

Please open an issue to discuss major changes before submitting a pull request.